Abstract The performance of Nb superconducting quantum devices is predominantly limited by dielectric loss at the metal–air interface, where Nb2O5 is considered the main loss source. Here, we suppress the formation of native oxides by in-situ deposition of a TiN capping layer on the Nb film. With TiN capping layers, no Nb2O5 forms on the surface of the Nb film. The quality factor Q i of the Nb resonator increases from 5.6 × 105 to 7.9 × 105 at low input power and from 6.8 × 106 to 1.1 × 107 at high input power. Furthermore, the TiN capping layer also shows good aging resistance in Nb resonator devices, with no significant performance fluctuations after one month of aging. These findings highlight the effectiveness of TiN capping layers in enhancing the performance and longevity of Nb superconducting quantum devices.
Abstract The performance of Nb superconducting quantum devices is predominantly limited by dielectric loss at the metal–air interface, where Nb2O5 is considered the main loss source. Here, we suppress the formation of native oxides by in-situ deposition of a TiN capping layer on the Nb film. With TiN capping layers, no Nb2O5 forms on the surface of the Nb film. The quality factor Q i of the Nb resonator increases from 5.6 × 105 to 7.9 × 105 at low input power and from 6.8 × 106 to 1.1 × 107 at high input power. Furthermore, the TiN capping layer also shows good aging resistance in Nb resonator devices, with no significant performance fluctuations after one month of aging. These findings highlight the effectiveness of TiN capping layers in enhancing the performance and longevity of Nb superconducting quantum devices.
Superconducting coplanar waveguide resonators are essential components in quantum processors, where their internal quality factor (Qi) constrains qubit coherence and readout fidelity. In niobium devices, microwave losses at millikelvin temperatures are strongly influenced by two-level systems (TLS) associated with the complex NbOx surface oxide. To mitigate these losses, we investigate a surface-engineering approach in which Nb films are capped in situ with a thin tantalum layer to suppress Nb2O5 formation and replace the native NbOx interface with a Ta-based oxide.
We fabricate Nb/Ta bilayer and reference Nb resonators on high-resistivity silicon using identical DC sputtering and wet etching conditions, and characterize their performance at millikelvin temperatures. Fresh Ta-encapsulated devices exhibit internal quality factors up to 2.4 x 10^6 in the near-single-photon regime, with power dependence consistent with reduced TLS-related loss at the metal-air interface. A control Nb device fabricated under the same process shows comparatively lower Q_TLS, consistent with the beneficial effect of the Ta capping layer. Furthermore, ageing tests performed on Nb/Ta resonators after six months reveal a moderate reduction in Q_TLS relative to their initial values, yet the performance remains superior to newly fabricated Nb-only devices. These results suggest that thin Ta encapsulation enhances interface quality and contributes to improved temporal stability while remaining compatible with Nb-based fabrication workflows.
The current performance of superconducting circuit-based quantum processors is limited by the poor understanding of interface physics, including native surface oxide formation on the superconducting metal, which causes two-level system (TLS) loss. Niobium (Nb), a superconducting metal with a high energy gap, is an ideal choice for superconducting processors, but unfortunately, it is marred by TLS. Several methods have been proposed to minimize surface oxide on the Nb film, and considerable improvement in TLS loss has been demonstrated. These methods include surface passivation through metal capping, self-assembly of organic molecules, and post-cleaning processes. Among these, metal capping is a suitable choice despite forming a 3–5 nm thick oxide, as self-assembly and post-treatment do not protect the Nb film surface during further fabrication. Here, we have proposed ruthenium (Ru) as a capping layer, forming a self-limiting oxidation with a 0.6 nm oxide thickness and predominantly producing fewer oxide compositions while being chemically resistant for further wafer fabrication processes, thus fulfilling all the criteria of an ideal capping layer. Our investigation suggests that Nb/Ru resonators have great potential as versatile and promising tools for advancing superconducting quantum technologies and integrating quantum interconnects into qubits with minimized TLS loss.
Niobium thin films are central to superconducting qubits, but their complex native oxides contribute significantly to microwave losses. A promising mitigation strategy is to suppress oxide formation using engineered thin-film encapsulation layers. Recent experiments have shown that ∼10 nm metallic overlayers on Nb capacitor films can improve the energy relaxation time T 1 of transmon qubits. Here, we present a comparative study of Au-capped Nb films fabricated in situ by molecular beam epitaxy and ex situ by sequential deposition benchmarked against bare Nb films. Using complementary structural, chemical, and spectroscopic techniques, we correlate the interface quality with superconducting electronic properties of the Au surface. Low-temperature scanning tunneling spectroscopy (STS) provides spatially resolved quasiparticle density-of-states maps. While both Au-capped Nb samples exhibit large areas with uniform, fully gapped density of states, clear differences emerge between the two interfaces. Compared to in situ Nb–Au, ex situ Nb–Au exhibits a reduced induced superconducting gap, broadened coherence peak, and localized in-gap states, consistent with a residual Nb x O y layer at the interface. Supported by complementary structural and chemical analyses, these findings demonstrate that STS directly links nanoscale superconducting properties to interface preparation, highlighting the importance of controlled in situ encapsulation for minimizing dissipation and improving quantum coherence.
Niobium and aluminum have been widely used to realize superconducting resonators and qubits. However, the native oxide at the metal air interface limits the devices results. This layer is highly disordered, and it induces losses limiting the performances of the resonators or qubits. Recently, tantalum led to great improvement in qubit relaxation time. This improvement has been attributed to a thinner and less disordered oxide layer compared to Nb and Al. This study is aimed to limit the formation of the oxide layer, by the introduction of novel superconducting materials such as rhenium or by encapsulating the Nb surface with other superconducting materials such as Ta, Al, Re or other non-superconducting materials such as Au. To better understand the results, we are realizing these studies on coplanar waveguide resonators where the relation between surface losses and quality factor is clear. We are also analyzing different fabrication techniques to completely characterize our devices, linking material properties to the performances. In this work we are also varying the participation ratio of the devices in order to extract with higher fidelity and precision the losses introduced by the involved interfaces.
To avoid degradation of circuit performance, fabrication processes for niobium-based superconductive electronics are typically limited to temperatures below 150°C. In this study, we investigated protective dielectric capping layers that preserve the superconducting properties of Niobium (Nb) wiring at processing temperatures as high as 400°C. To assess the thermal stability of Nb films, 400 nm thick Nb layers were deposited on oxidized silicon wafers and were either left uncapped or were intentionally capped with selected dielectric materials. The samples were subjected to post-deposition annealing in an argon atmosphere for comparison. Samples were annealed up to 450°C, and changes in room temperature sheet resistance were used as a proxy for film degradation and checked with cryogenic measurements of samples annealed up to 400°C. A sharp increase in room temperature sheet resistance, and a resulting decrease in residual resistivity ratio was observed in pristine Nb films above 300°C. We found that theex-situdeposition of insulating covering (“cap”) reduces the onset of degradation. Notably, silicon oxide (SiO_x) capping shows significant improvement and silicon nitride (Si_xN_y) capped samples exhibit minimal changes in resistance across the full annealing range. These results suggest that Si_xN_yencapsulation as a dielectric “cap” preserves superconductive properties and offers an expanded thermal budget for Nb-based superconductive electronics, with implications for multilayer integration and scalable fabrication.
Substrate surface preparation is a key step in the fabrication of electronic devices. In III--V semiconductor platforms, e.g. used in HEMTs and lasers, removal of the native substrate oxide is extremely important, where improper removal will negatively affect end-of-line device performance. However, the impact of substrate preparation in hybrid superconductor--semiconductor (S--Sm) systems relevant to quantum information applications remains poorly understood. This study compares three surface preparations for the deposition of sputtered Nb films on InP: (i) no intentional oxide removal (control), (ii) \textit{in-situ} Ar+ milling, and (iii) S-passivation. \textit{In-situ} Ar+ milling reduces the O concentration at the metal--substrate (MS) interface, but also roughens the InP surface, increasing the effective thickness of the Nb--InP interface and promoting O incorporation through extended defects in the Nb film. S-passivation suppresses interfacial O more effectively while preserving a sharper and smoother buried interface. It also yields Nb films with higher superconducting transition temperatures and less structural damage than the Ar+-milled samples. Despite these materials improvements, the microwave response is comparable across the three preparations. At single photon powers, the highest internal quality factors, Qi, are approximately 1.30×105 (130k) for the control resonators, 9.7×104 (97k) for the S-passivated resonators, and 8.4×104 (84k) for the Ar+-milled resonators. These results suggest that the present devices are not primarily limited by dielectric loss at the buried Nb--InP interface.